[AArch64][GlobalISel] Custom selection for s8 load acquire.

Implement this single atomic load instruction so that we can compile stack
protector code.

Differential Revision: https://reviews.llvm.org/D66245

llvm-svn: 368923

lib/Target/AArch64/AArch64InstructionSelector.cpp

@@ -1740,7 +1740,14 @@ bool AArch64InstructionSelector::select(MachineInstr &I) {
 
     auto &MemOp = **I.memoperands_begin();
     if (MemOp.getOrdering() != AtomicOrdering::NotAtomic) {
-      LLVM_DEBUG(dbgs() << "Atomic load/store not supported yet\n");
+      // For now we just support s8 acquire loads to be able to compile stack
+      // protector code.
+      if (MemOp.getOrdering() == AtomicOrdering::Acquire &&
+          MemOp.getSize() == 1) {
+        I.setDesc(TII.get(AArch64::LDARB));
+        return constrainSelectedInstRegOperands(I, TII, TRI, RBI);
+      }
+      LLVM_DEBUG(dbgs() << "Atomic load/store not fully supported yet\n");
       return false;
     }
     unsigned MemSizeInBits = MemOp.getSize() * 8;

test/CodeGen/AArch64/GlobalISel/select-atomic-load-store.mir

@@ -0,0 +1,37 @@
+# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
+# RUN: llc -mtriple=aarch64-- -run-pass=instruction-select -verify-machineinstrs %s -o - | FileCheck %s
+--- |
+  target datalayout = "e-m:e-i8:8:32-i16:16:32-i64:64-i128:128-n32:64-S128"
+  target triple = "aarch64"
+
+  define i8 @load_acq_i8(i8* %ptr) {
+    %v = load atomic i8, i8* %ptr acquire, align 8
+    ret i8 %v
+  }
+
+...
+---
+name:            load_acq_i8
+alignment:       2
+legalized:       true
+regBankSelected: true
+tracksRegLiveness: true
+liveins:
+  - { reg: '$x0' }
+machineFunctionInfo: {}
+body:             |
+  bb.1:
+    liveins: $x0
+
+    ; CHECK-LABEL: name: load_acq_i8
+    ; CHECK: liveins: $x0
+    ; CHECK: [[COPY:%[0-9]+]]:gpr64sp = COPY $x0
+    ; CHECK: [[LDARB:%[0-9]+]]:gpr32 = LDARB [[COPY]] :: (load acquire 1 from %ir.ptr, align 8)
+    ; CHECK: $w0 = COPY [[LDARB]]
+    ; CHECK: RET_ReallyLR implicit $w0
+    %0:gpr(p0) = COPY $x0
+    %2:gpr(s32) = G_LOAD %0(p0) :: (load acquire 1 from %ir.ptr, align 8)
+    $w0 = COPY %2(s32)
+    RET_ReallyLR implicit $w0
+
+...